Anyone building race engines has to deal with the question of how thin to make the vertical squish clearance between piston and head. "Squish" refers to those areas of the piston that come very close to the head at top dead center (TDC) in order to squeeze out the fuel-air mixture from in between, forming jets that give the charge a stir to accelerate combustion.

With that in mind, sometime in the late 1970s, I assembled one Yamaha TZ250 engine at a squish clearance of 0.018 inch rather than the usual 0.028. After a couple of practices, I drained the cooling water and pulled the cylinder head. Aha! There were bright places all over the squish zones, indicating that kissing contact had occurred.

When you measure squish clearance, you symmetrically place bits of something soft, solder or modeling clay, on top of the piston and then turn the engine past TDC to compress it. Then you measure its thickness.

This takes up all the bearing clearances in a downward direction—piston to wrist pin to connecting rod small end; con-rod big end to crankpin; and crank main bearings to crankcase. What if we put a dial gauge against the piston crown and then lift up on the piston, taking up all the clearance in the opposite direction?

Just now, I ran up to the shop, pulled out an old “checking” engine (a set of two-stroke crankcases with a crank, piston, and cylinder for checking timings and port heights), and I tried it. With the piston at TDC and the dial gauge bearing on its crown, I first pushed down on the piston, noted the reading, then with a finger in the intake port, I pushed up on a wrist-pin boss. The gauge showed roughly 0.001 inch of movement.

So how could the piston be just touching the head with a squish clearance of 0.018 inch?

John Kocinski
Yamaha 250cc two-stroke production racers were long powered by parallel twins. That gave way in 1991 to a 90-degree V-twin based on the YZR250 factory machines, like the 0WB9, which carried American John Kocinski to seven wins and the 1990 250cc world title.Courtesy of Yamaha

Because, at peak revs of 11,000, the piston is reaching a peak deceleration of 4,500 G as it reaches TDC, and that bends and stretches the moving parts. According to my little scale, piston, ring, wrist pin, and wrist-pin bearing weigh 174 grams and the small end of the rod 72 grams, so the tension in the rod required to reverse their motion at TDC and 11,000 rpm was 2,700 pounds. The rod was stretching, the wrist pin bending, the piston’s wrist-pin bosses bending as well. In the top and bottom con-rod needle bearings, the cylindrical needles were flattening slightly at their line contacts with their races, which, in turn, were being elastically “dented” by that same force. Down below, the crankpin was being pulled upward, bending the crank. And the crank main bearings—one ball and one roller bearing—were likewise deflecting.

Anyone who has rebuilt these pressed-together cranks can see proof that they flex from the reddish discoloration that forms in the press fits, evidence of ever-so-slight back-and-forth relative motion.

One MotoGP manufacturer in the early years of the series had shifting problems that seemed to come from bind caused by crankcase flexure.

Preparing a new bike at the start of the season taught us that if there wasn’t a half-inch clearance between the front tire (with suspension fully compressed) and the face of the coolant radiator, there would be contact during hard braking. The fork tubes were bending, the steering head of the steel tube chassis was deflecting, and the bias front tire was growing in diameter. And at the next corner after a bumpy corner exit, the brake lever might come straight to the bar. Passing over bumps at lean angle kicks the front wheel out of plane by flexing the front axle, allowing the brake disc(s) to tilt enough to knock the brake pads back in their bores. Hope there’s time for a quick second pull…

Whole engines flex when used as structural chassis members. One MotoGP manufacturer in the early years of the series had shifting problems that seemed to come from bind caused by crankcase flexure. Another team found that force from the rear suspension unit, attached at its top to a rear chassis crossmember, was bending it enough, as if it were a bowstring, to pull the chassis uprights inward against the swingarm at its pivot, causing suspension bind and weird handling. The attachment point was moved to a heavy lug machined integral with the gearbox.

Everything is flexible because atoms in solids are bound to each other by nothing but springy invisible electric fields. Careful design takes such materials’ flex into consideration, keeping pistons from tapping against cylinder heads and front tires from rubbing through radiator tubes. In the past, this was done by experience or by manual calculation. Today, it can be predicted by computer analysis, but someone has to remember to actually do it and use the result to prevent bad stuff from happening.